1. Field of the Invention
The present invention relates to a temporary center system in a distributed data base system consisting of a plurality of terminal units connected in the network system.
In these years, a distributed processing system or distributed data base system (hereinafter referred to as distributed data base system) has been structured with a plurality of terminal units connected to the public or private network. In such a system, the data bases provided in each terminal unit are distributed in many areas coupled through a real time network (communication network) and the capability for processing such data bases is also distributed.
The distributed data base system naturally provides a greater part of necessary data within the data base of self terminal units. However, if the necessary data is provided in other terminal units, such data must be received through the network. Therefore, in this case, the distributed data base system is required to improve the disadvantage that a longer period of time is taken for transfer of data.
2. Background of the Invention
A data base system has been structured in these years with the terminal units, which are connected in the network, and provide a plurality of information processing functions. From the view point of control, the conventional data base systems may be classified into the centralized type system and distributed type data base system. The centralized data base system controls all data bases with the one center host and retrieves necessary information by accessing the center host from the other terminal units.
This system has a problem in that the center host becomes large in scale and the load of the center host becomes excessive because accesses from the other terminal units to the center host are centralized. As a system to solve such a problem, a distributed type data base system has been used. In the case of this data base system, a plurality of terminal units connected in the network system is respectively provided with data base to be controlled by each terminal unit.
A conventional distributed data base System will be explained with reference to FIG. 16. In this figure, T1-T4 denote terminal units, T1-T4 respectively providing information processing means (CPU, memory, input/output unit, file unit, etc.). F denotes a file for storing data bases stored in each terminal unit. NW denotes a network (public network or private network).
FIG. 16A indicates the basic structure of a distributed data base system. A plurality of terminal units T1-T4 provides a greater part of necessary data within the data base thereof. However, in case the data base is provided in other terminal unit (in case it is detected the related data is provided in the other terminal unit), the terminal unit requests the data by sequentially making inquiries to the other terminal units through the network. For instance, when the terminal unit T1 requests data from terminal unit T2, the data is extracted from a file unit in the terminal unit T2 when such data exists therein, such data is then transferred to the terminal unit T1 through the network.
In the conventional distributed type data base system shown in FIG. 16A, in case each terminal unit does not have the data of the other terminal units in the data base of the self terminal unit, when each terminal unit cannot find the necessary data by retrieving the data base of the self terminal unit, each terminal unit sequentially issues an inquiry to the other terminal units whether there is the necessary data or not through the network system. When a certain other terminal unit has such data, each terminal unit receives the necessary data from such other terminal unit through the network. Thereby, there is a problem in that the load applied on the terminal units and network becomes excessive.
Moreover, in case each terminal unit always holds all data in the system, this system has a problem in that when data is generated, such data must be transferred to each terminal unit and thereby a load on the network becomes excessive like the preceding case.
In addition, when the data transmitted to each terminal is revised, the revised data (revised reference) must be transmitted to each terminal (this is called the missing operation).
The distributed type data base system has a problem, as explained above, in that each terminal unit operates taking a longer period to retrieve and gather necessary data, the application period of network increases and the load of terminal units and the load of network become remarkable in comparison with the number of terminal units.
On the other hand, in place of the direct transfer of data, a command transfer system has been proposed. In this case, data is expressed by compressed command (code) and a quantity of information at the time of transfer is decreased. The structure of this system is shown in FIG. 16B.
The command transfer system will be explained. For example, in the terminal unit T2, "VOL1" is assigned as the compressed command to certain data. Here, it is assumed that the terminal unit T1 issues the data transfer request to the terminal unit T2.
In this case, the terminal unit T2 does not transfer the data but transfers the assigned command "VOL1" and the number information "T2" as the additional information, which indicates the terminal unit has such data. Upon reception of the command, the terminal unit T1 retrieves data corresponding to the command and stores the retrieved data in the self (T1) data base.
In this case, if there is no data corresponding to the command in the self data base, the terminal unit T1 makes access to the terminal unit T2 using the number information "T2" of the terminal unit added to the command and requests the data corresponding to the command "VOL1". Thereby, the terminal unit T1 acquires the object data.
In the same way, when the command "VOL2" and "T3", which indicates the position of data, are transmitted from the terminal unit T3 responding to the data request from the terminal unit T1, such data may be obtained in the same sequence as that explained above.
In some cases, a data group consisting of a plurality of data or a plurality of integrated data is processed in relation with each other in the data base. However, the data base utilizing the commands shown in FIG. 16B has a problem in that the load on the network and terminal units is increased and a longer time is also taken because inquiry is sequentially made to each terminal unit in order to confirm whether there is related data or not. Then, the data is gathered and the editing process is carried out.
In addition, when a request to read the data in the data base or a request to obtain the data is issued in the data base system, the qualification for reading the data is checked, except for the data to be opened to any terminal unit (or person). In the conventional system, it is checked whether a user has the reasonable qualification or not for the data read request (data request) issued from the user (or the terminal unit) at each terminal unit. This check is usually conducted using a personal ID (ID number), pass word, or keyword. Namely, the personal ID precedingly registered as the qualification is collated with the personal ID input by user. When these IDs coincide, data reading is allowed. However, according to this conventional reading system, even if a person does not have reasonable qualification, he can read the data base any time he wants only by inputting the reasonable personal ID.
Therefore, there is a problem in that any information may be read with any personal ID without relation to importance of data. This problem is particularly important in the distributed type data base connected to the public network. Namely, it is desirable that the data other than that which should be kept as secret (closed data) should be read (opened) from any terminal units. Thereby, it becomes possible to read the data with each other between the physically separated terminal units and to realize the processing using such data. Therefore, this system can fully characterize the connection with the public network and the features of the distributed data base system.
But, on the other hand, this system cannot satisfy the request that the data should naturally be read (should be set as the open data) between different terminal units only by the person having the qualification such as the personal ID. Namely, the conventional system has a problem in that the data base has been closed because it is impossible to give a limitation on the qualification in accordance with the importance of data.